[entrep]

> - But if the malicious actor owns 2/3 of validators, what does he care what other nodes do ? They effectively control what block is accepted

As I understand it, anyone who owns that amount of ETH would probably be interested in contributing to the trust of the Ethereum blockchain.

[falcolas]

What reason would anyone have to believe in their benevolence? It would be safer to believe in and use their greed and self preservation.

[mastazi]

If you own a large amount of a currency, you want to ensure that the currency in question is trusted, or otherwise that currency would end up losing its value. It is your greed that is guiding you, not your benevolence.

[troutwine]

Eh, I think it’s naive to believe that greed is sufficient protection. Owning a bunch of ETH so you can set ETH on fire would be a waste of money, only so long as the person setting the fire didn’t value the mischief more than the ETH. I don’t see why nation states wouldn’t attack each other by burning down the crypto assets of citizens, either.

[mastazi]

> I think it’s naive to believe that greed is sufficient protection.

I never said that, I actually agree with you. I was just explaining GP's point to parent who in my opinion missed it.

[troutwine]

Ah, apologies. I read you as agreeing with and clarifying GP's argument.

[dmitriid]

> If you own a large amount of a currency,

you rug pull, divest into hard cash at peak, <any of the dozens of ways people with large amounts of cryptocurrencies get rid of those>

[falcolas]

One problem with this - speculative assets are at least partially valuable because of their volatility - their ability to appreciate quickly. Even if eth is untrusted for long term value, so long as it has the potential to make money for someone in the short term it's going to continue to get transactions.

And, as always, sometimes folks just want to watch the world burn. It's a relatively low investment (~$50k) to become a validator, and the fine's only about $7k (.5 eth).

Imagine the damage someone with a few million (say, some baby-boomer who just sold their California duplex) could do if they wanted to.

[lottin]

Google short selling.

[mastazi]

thanks, I work at a financial institution, so I "Googled" it a few times, short selling is based on borrowing the underlying securities - not owning them - or alternatively owning a(n often bespoke) derivative, related to those securities, which behaves as a short.

If you own a security you are long, not short.

[johnthewise]

If you hold eth, you are long eth.

[lottin]

Not really, you can be long in one market, and short in the derivatives market, and your net position be short. You'd be holding eth, but you'd be short.

[mastazi]

not really, in that case you have two separate positions, one long and one short.

[WFHRenaissance]

No one would use a network with this level of centralization, or at least, it would be unwise to.

[cowtools]

like a 51% attack on a PoW through pool conspiracy, it will be impossible to tell if (when?) the network is controlled by a single entity.

By the time the public has figured out the plot, it's too late: the attacker has already run to the exchanges with double-spent coins.

[rakoo]

Call me a pessimist, but I have stopped assuming rationality of all actors acting to secure a long-term future, especially in the world of cryptocurrencies

[pclmulqdq]

PoW is magical in that these actors are strongly incentivized to secure the network. PoS is not the same.

[rakoo]

This was the whole point of blockchains. No one trusts anyone, everyone is in it for themselves only. The Adam Smith model. And it works, if you don't loom at the disastrous environmental impact.

[ww520]

Fund can be borrowed. As attested in last the couple months, billions of funds were borrowed to bet on the market one way or the other. Borrowing a billion dollars to subvert the network for a short duration sounds like a feasible strategy.

[stepanhruda]

You can’t borrow enough money and immediately take over the network. Those funds need to be staked first in validators with 32 ETH each and you need to control enough validators to control consensus. It’s more difficult than PoW in a way since you can’t just start validating, there is an entry/exit queue.

[lottin]

This isn't argument against PoS, because mining rigs can also be borrowed. Anything can be borrowed.

[boosteri]

Why would it need to be a significant amount of ETH? Could someone coordinate an attack on validator nodes, making themselves 67%+ temporarily?

[stepanhruda]

Yes. That’s why diversity of client implementations is important.

[yunohn]

Sure, but that wasn’t the question. How does the punishment play out, assuming they’re acting in bad faith?

[rglullis]

If you have 2/3 of the validators, there is no way that you can be punished. You will effectively be able to produce the blocks and tell the network that they are valid.

If you don't have that, the most that you can do is to attack the network by proposing bad blocks to slow down block production. Anytime that a (selected) validator proposes a bad block, the other validators that catch your mistake will snitch on you (through attestations) and the validator will get their stake funds slashed. The more your funds are slashed, the less of a chance you will have to be proposing new blocks in the future.